| |
BGonline.org Forums
Regarding market losers
Posted By: Chuck Bower In Response To: Regarding market losers (Bob Koca)
Date: Tuesday, 10 March 2026, at 10:21 p.m.
I don't understand why you chose this position as an illustration of Blue's strength. White being on roll (which occurs after {21,22,23,24} as here) is quite valuable. BTW, all of my conclusions below are based upon 4-ply and XGR+ rollouts, not evaluations or mini-rollouts.
I looked deeper into this position but with White on roll. The only White roll that I could find that 4-ply rollouts indicate as a pass is 31. The other 34 rolls are at least Takes (don't know if they are doubles or not -- didn't look at that).
I then looked at Blue's (apparently) better entering rolls {16,25,26}. For the first four of these, I could only find four White rolls that were passes: {61,62}. Blue's 62 from the bar is much better for him than any other roll. It appears there are a lot of passes after White's subsequent roll and I didn't feel like digging into those. However, I asked the question: "suppose after Blue's 62, White stops and wants to settle. What is the Fair Settlement Value (FSV) for this -- after Blue's best entering roll -- again assuming Blue owns the 2-cube?"
I teed up XG's strongest rollout settings which will run on my computer in a reasonable time -- XGR+ rollout. The result for 1296 trials was 0.92 +/- 0.009 (95% confidence, unit cube) equity for Blue. Let's assume this is being played for $50/pt. I conclude an FSV of 0.92*2*$50 = $92 to Blue, so an (upcoming) Take for White if averaged over all of her 36 rolls this turn.
FWIW, I don't think XG 3-ply is particularly good for a rollout here. That's why I used both 4-ply and XGR+ rollout settings, and even those might not tell the true story. To add fuel to that claim, XG++ (evaluation/mini-rollout) missed by a lot when evaluating various White checker plays after Blue's 2x entry rolls. "A lot" was as many as 0.100 compared to rollouts.
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.